| { | |
| "best_metric": 0.7326732673267327, | |
| "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-4/checkpoint-96", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 432, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2733114957809448, | |
| "learning_rate": 2.2702186710865246e-07, | |
| "loss": 0.7025, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.243804931640625, | |
| "learning_rate": 4.5404373421730493e-07, | |
| "loss": 0.6974, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.7711552381515503, | |
| "learning_rate": 6.810656013259573e-07, | |
| "loss": 0.696, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.1453403234481812, | |
| "learning_rate": 9.080874684346099e-07, | |
| "loss": 0.6989, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.2729355096817017, | |
| "learning_rate": 1.1351093355432624e-06, | |
| "loss": 0.6968, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 1.1592165231704712, | |
| "learning_rate": 1.3621312026519146e-06, | |
| "loss": 0.6959, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.1798148155212402, | |
| "learning_rate": 1.589153069760567e-06, | |
| "loss": 0.6952, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 2.1216671466827393, | |
| "learning_rate": 1.8161749368692197e-06, | |
| "loss": 0.6886, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.3416370153427124, | |
| "learning_rate": 2.043196803977872e-06, | |
| "loss": 0.6864, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_f1": 0.72, | |
| "eval_loss": 0.688262939453125, | |
| "eval_runtime": 1.3468, | |
| "eval_samples_per_second": 47.521, | |
| "eval_steps_per_second": 5.94, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.1856281757354736, | |
| "learning_rate": 2.169320063482679e-06, | |
| "loss": 0.6917, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 1.4077153205871582, | |
| "learning_rate": 2.1440954115817176e-06, | |
| "loss": 0.6884, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.1792664527893066, | |
| "learning_rate": 2.1188707596807562e-06, | |
| "loss": 0.6668, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.0386197566986084, | |
| "learning_rate": 2.093646107779795e-06, | |
| "loss": 0.6694, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 2.0565919876098633, | |
| "learning_rate": 2.0684214558788335e-06, | |
| "loss": 0.6561, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.2978509664535522, | |
| "learning_rate": 2.043196803977872e-06, | |
| "loss": 0.6789, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 2.058328628540039, | |
| "learning_rate": 2.0179721520769108e-06, | |
| "loss": 0.6633, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.6023226976394653, | |
| "learning_rate": 1.9927475001759494e-06, | |
| "loss": 0.6655, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.5510762929916382, | |
| "learning_rate": 1.967522848274988e-06, | |
| "loss": 0.6622, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 1.098602533340454, | |
| "learning_rate": 1.9422981963740267e-06, | |
| "loss": 0.6633, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6816024780273438, | |
| "eval_runtime": 1.3765, | |
| "eval_samples_per_second": 46.493, | |
| "eval_steps_per_second": 5.812, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.9589098691940308, | |
| "learning_rate": 1.9170735444730654e-06, | |
| "loss": 0.659, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 1.070695161819458, | |
| "learning_rate": 1.8918488925721038e-06, | |
| "loss": 0.6313, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.9913639426231384, | |
| "learning_rate": 1.8666242406711424e-06, | |
| "loss": 0.6652, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.0632878541946411, | |
| "learning_rate": 1.841399588770181e-06, | |
| "loss": 0.673, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.1036579608917236, | |
| "learning_rate": 1.8161749368692197e-06, | |
| "loss": 0.6451, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.08384108543396, | |
| "learning_rate": 1.7909502849682583e-06, | |
| "loss": 0.6322, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 0.9407000541687012, | |
| "learning_rate": 1.765725633067297e-06, | |
| "loss": 0.6755, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.9016568660736084, | |
| "learning_rate": 1.7405009811663356e-06, | |
| "loss": 0.5985, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 1.1134448051452637, | |
| "learning_rate": 1.7152763292653743e-06, | |
| "loss": 0.603, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6800689697265625, | |
| "eval_runtime": 1.3861, | |
| "eval_samples_per_second": 46.173, | |
| "eval_steps_per_second": 5.772, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 0.7627719640731812, | |
| "learning_rate": 1.6900516773644127e-06, | |
| "loss": 0.6557, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 0.9291415214538574, | |
| "learning_rate": 1.6648270254634511e-06, | |
| "loss": 0.6219, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 0.9248765707015991, | |
| "learning_rate": 1.6396023735624898e-06, | |
| "loss": 0.6325, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 0.9842573404312134, | |
| "learning_rate": 1.6143777216615284e-06, | |
| "loss": 0.6521, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 0.8689214587211609, | |
| "learning_rate": 1.589153069760567e-06, | |
| "loss": 0.5929, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 1.0012000799179077, | |
| "learning_rate": 1.5639284178596057e-06, | |
| "loss": 0.584, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.7438368797302246, | |
| "learning_rate": 1.5387037659586443e-06, | |
| "loss": 0.6813, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 1.8603870868682861, | |
| "learning_rate": 1.513479114057683e-06, | |
| "loss": 0.6099, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 0.9918416738510132, | |
| "learning_rate": 1.4882544621567216e-06, | |
| "loss": 0.6192, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 1.9146322011947632, | |
| "learning_rate": 1.4630298102557603e-06, | |
| "loss": 0.6472, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6818161010742188, | |
| "eval_runtime": 1.3841, | |
| "eval_samples_per_second": 46.239, | |
| "eval_steps_per_second": 5.78, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 0.9502781629562378, | |
| "learning_rate": 1.437805158354799e-06, | |
| "loss": 0.6447, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.8570067286491394, | |
| "learning_rate": 1.4125805064538375e-06, | |
| "loss": 0.5306, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 0.8097484111785889, | |
| "learning_rate": 1.3873558545528762e-06, | |
| "loss": 0.6202, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 2.0106472969055176, | |
| "learning_rate": 1.3621312026519146e-06, | |
| "loss": 0.6705, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 1.090775489807129, | |
| "learning_rate": 1.3369065507509533e-06, | |
| "loss": 0.6297, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 0.8988145589828491, | |
| "learning_rate": 1.311681898849992e-06, | |
| "loss": 0.5896, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 0.9149978756904602, | |
| "learning_rate": 1.2864572469490305e-06, | |
| "loss": 0.6156, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 1.9398412704467773, | |
| "learning_rate": 1.2612325950480692e-06, | |
| "loss": 0.6305, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.9217966794967651, | |
| "learning_rate": 1.2360079431471078e-06, | |
| "loss": 0.5943, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.9083653688430786, | |
| "learning_rate": 1.2107832912461465e-06, | |
| "loss": 0.6386, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6846389770507812, | |
| "eval_runtime": 1.4094, | |
| "eval_samples_per_second": 45.409, | |
| "eval_steps_per_second": 5.676, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.9323675036430359, | |
| "learning_rate": 1.1855586393451851e-06, | |
| "loss": 0.5779, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 0.7549787163734436, | |
| "learning_rate": 1.1603339874442238e-06, | |
| "loss": 0.5948, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "grad_norm": 0.8535837531089783, | |
| "learning_rate": 1.1351093355432624e-06, | |
| "loss": 0.6928, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "grad_norm": 1.2038137912750244, | |
| "learning_rate": 1.109884683642301e-06, | |
| "loss": 0.5887, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 0.9501279592514038, | |
| "learning_rate": 1.0846600317413395e-06, | |
| "loss": 0.5776, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "grad_norm": 0.7421719431877136, | |
| "learning_rate": 1.0594353798403781e-06, | |
| "loss": 0.6734, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 0.8555863499641418, | |
| "learning_rate": 1.0342107279394168e-06, | |
| "loss": 0.6399, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 0.8841156363487244, | |
| "learning_rate": 1.0089860760384554e-06, | |
| "loss": 0.6173, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "grad_norm": 0.8565478324890137, | |
| "learning_rate": 9.83761424137494e-07, | |
| "loss": 0.5537, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6864242553710938, | |
| "eval_runtime": 1.3622, | |
| "eval_samples_per_second": 46.983, | |
| "eval_steps_per_second": 5.873, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 0.8750139474868774, | |
| "learning_rate": 9.585367722365327e-07, | |
| "loss": 0.5531, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "grad_norm": 1.0445302724838257, | |
| "learning_rate": 9.333121203355712e-07, | |
| "loss": 0.638, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 0.7958914637565613, | |
| "learning_rate": 9.080874684346099e-07, | |
| "loss": 0.547, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "grad_norm": 0.9992254376411438, | |
| "learning_rate": 8.828628165336485e-07, | |
| "loss": 0.6425, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 6.46, | |
| "grad_norm": 0.8400682806968689, | |
| "learning_rate": 8.576381646326871e-07, | |
| "loss": 0.6955, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "grad_norm": 0.742438793182373, | |
| "learning_rate": 8.324135127317256e-07, | |
| "loss": 0.6473, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 0.6693254113197327, | |
| "learning_rate": 8.071888608307642e-07, | |
| "loss": 0.603, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.77, | |
| "grad_norm": 1.0816401243209839, | |
| "learning_rate": 7.819642089298028e-07, | |
| "loss": 0.6053, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "grad_norm": 0.7275277376174927, | |
| "learning_rate": 7.567395570288415e-07, | |
| "loss": 0.612, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 6.98, | |
| "grad_norm": 0.7834873795509338, | |
| "learning_rate": 7.315149051278801e-07, | |
| "loss": 0.55, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6889228820800781, | |
| "eval_runtime": 1.3769, | |
| "eval_samples_per_second": 46.483, | |
| "eval_steps_per_second": 5.81, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 7.08, | |
| "grad_norm": 1.24147367477417, | |
| "learning_rate": 7.062902532269188e-07, | |
| "loss": 0.508, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "grad_norm": 1.8932181596755981, | |
| "learning_rate": 6.810656013259573e-07, | |
| "loss": 0.6358, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "grad_norm": 1.861436128616333, | |
| "learning_rate": 6.55840949424996e-07, | |
| "loss": 0.5741, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.8429200053215027, | |
| "learning_rate": 6.306162975240346e-07, | |
| "loss": 0.5717, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.8665741682052612, | |
| "learning_rate": 6.053916456230732e-07, | |
| "loss": 0.6992, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 2.312748908996582, | |
| "learning_rate": 5.801669937221119e-07, | |
| "loss": 0.6151, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 7.71, | |
| "grad_norm": 1.1628329753875732, | |
| "learning_rate": 5.549423418211505e-07, | |
| "loss": 0.5354, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 7.81, | |
| "grad_norm": 1.8674992322921753, | |
| "learning_rate": 5.297176899201891e-07, | |
| "loss": 0.6411, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "grad_norm": 0.7112137675285339, | |
| "learning_rate": 5.044930380192277e-07, | |
| "loss": 0.6063, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6909217834472656, | |
| "eval_runtime": 1.3913, | |
| "eval_samples_per_second": 45.999, | |
| "eval_steps_per_second": 5.75, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 1.201416015625, | |
| "learning_rate": 4.792683861182663e-07, | |
| "loss": 0.647, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "grad_norm": 0.9348795413970947, | |
| "learning_rate": 4.5404373421730493e-07, | |
| "loss": 0.5356, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 8.23, | |
| "grad_norm": 1.895739197731018, | |
| "learning_rate": 4.2881908231634357e-07, | |
| "loss": 0.5384, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "grad_norm": 1.8363467454910278, | |
| "learning_rate": 4.035944304153821e-07, | |
| "loss": 0.541, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 8.44, | |
| "grad_norm": 2.475804090499878, | |
| "learning_rate": 3.7836977851442075e-07, | |
| "loss": 0.5503, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "grad_norm": 1.314663290977478, | |
| "learning_rate": 3.531451266134594e-07, | |
| "loss": 0.575, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "grad_norm": 1.850918173789978, | |
| "learning_rate": 3.27920474712498e-07, | |
| "loss": 0.6658, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "grad_norm": 0.7412477135658264, | |
| "learning_rate": 3.026958228115366e-07, | |
| "loss": 0.5632, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 8.85, | |
| "grad_norm": 1.3282320499420166, | |
| "learning_rate": 2.7747117091057526e-07, | |
| "loss": 0.7303, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "grad_norm": 1.019906759262085, | |
| "learning_rate": 2.5224651900961385e-07, | |
| "loss": 0.7438, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6923065185546875, | |
| "eval_runtime": 1.3679, | |
| "eval_samples_per_second": 46.786, | |
| "eval_steps_per_second": 5.848, | |
| "step": 432 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 480, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 6549027725117760.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 2.1794099242430636e-06, | |
| "per_device_train_batch_size": 4 | |
| } | |
| } | |